Although acute myeloid leukemia (AML) causes substantial treatment burden in both children and adults, AML treatment response remains incompletely understood. This application proposes a genome-wide association study (GWAS) of pediatric AML treatment response. Two hypotheses underlie this application. First, somatic genetic variation will modify AML relapse and treatment related infection risk. Second, clinical trial simulations (CTS) may define how genetic variation data may be used to modify AML treatment. These hypotheses will be tested in three specific aims. Aim 1 will use the Illumina Human 610 HH Bead Chip to perform a GWAS with 840 patients to identify genotypes associated with AML relapse and infectious complication risk. Aim 2 will use the Illumina Infinium Chip to validate 4% of SNPs found in Aim 1 in 1,160 Caucasian patients and 750 non-Caucasian patients. Aim 3 will use CTS to test the clinical applicability of infection susceptibility genotypes discovered and validated in the first two aims. This application has significance both as the first GWAS in AML and as a novel step towards translating genetic variability data into the clinical care of patients. Furthermore, since pediatric and adult AML share common molecular pathologies and treatment strategies, this research may inform the care of both adult and pediatric AML patients.